The Systematic Properties of the Warm Phase of Starburst-Driven Galactic Winds

Abstract

Using ultra-violet absorption-lines, we analyze the systematic properties of the warm ionized phase of starburst-driven winds in a sample of 39 low-redshift objects that spans broad ranges in starburst and galaxy properties. Total column densities for the outflows are ∼10 21 cm −2 . The outflow velocity (v out ) correlates only weakly with the galaxy stellar mass (M * ), or circular velocity (v cir ), but strongly with both SFR and SFR/area. The normalized outflow velocity (v out /v cir ) correlates well with both SFR/area and SFR/M * . The estimated outflow rates of warm ionized gas (Ṁ ) are ∼ 1 to 4 times the SFR, and the ratioṀ /SF R does not correlate with v out .We show that a model of a population of clouds accelerated by the combined forces of gravity and the momentum flux from the starburst matches the data. We find a threshold value for the ratio of the momentum flux supplied by the starburst to the critical momentum flux needed for the wind to overcome gravity acting on the clouds (R crit ). For R crit > 10 (strong-outflows) the outflows momentum flux is similar to the total momentum flux from the starburst and the outflow velocity exceeds the galaxy escape velocity. Neither is the case for the weak-outflows (R crit < 10). For the weak-outflows, the data severely disagree with many prescriptions in numerical simulations or semianalytic models of galaxy evolution. The agreement is better for the strong-outflows, and we advocate the use of R crit to guide future prescriptions.

NEAR-ULTRAVIOLET SPECTROSCOPY OF STAR-FORMING GALAXIES FROM eBOSS: SIGNATURES OF UBIQUITOUS GALACTIC-SCALE OUTFLOWS

“…The outflow mechanisms can not only expel gas out of galaxies, but can also result in a greatly reduced rate (quenching) in star formation, which is a necessary component in any modern galaxy formation theory in order to match the observed stellar mass function of galaxies and theoretical mass function of dark matter halos and explain the observed bimodality of star-forming and quiescent galaxies (e.g., Croton n et al 2006;Somerville et al 2008). Our observations provided strong evidence for ubiquitous galactic-scale outflows driven by star formation as one of the most important gas sources, in agreement with previous works (e.g., Heckman et al 1990Heckman et al , 2000Heckman et al , 2015Steidel et al 1996Steidel et al , 2010Pettini et al 2001;Shapley et al 2003;Martin 2005;Rupke et al 2005;Veilleux et al 2005;Weiner et al 2009, among others). The exact mechanisms driving the outflows are still a debated topic, with propositions including ram pressure from hot outflowing gas from supernovae (e.g., Chevalier & Clegg 1985;Murray et al 2005) and radiation pressure from massive star clusters (e.g., Nath & Silk 2009;Murray et al 2011).…”

Molecular Outflows in Local ULIRGs: Energetics from Multitransition OH Analysis

“…Starburst99 models (Leitherer et al 1999) predict that a 40 Myr old continuous starburst can supply a maximum momentum (see also Veilleux et al 2005;Heckman et al 2015) ofṖ * ∼ 3.5L * /c (including radiation pressure at a level of L * /c), which is marked as a shaded rectangle in Fig. 18.…”

“…In this paper, we use the most conservativeṀ out ,Ṗ out , andĖ out values to characterize the outflows. These are the "timeaveraged thin shell" values in Rupke et al (2005c), which have also been used by a number of authors describing the energetics of the ionized and neutral phases of outflows (e.g., Rupke & Veilleux 2013a;Arav et al 2013;Borguet et al 2013;Heckman et al 2015), and are most appropriate for comparison with outflow models (e.g., Faucher-Giguère & Quataert 2012; Stern et al 2016;Thompson et al 2015).…”

“…Although τ IR is expected to vary with r, we avoid overestimating its effect (see Roth et al 2012) by simply taking a characteristic constant value of µm H k IR N H , where an absorption-mass coefficient of k IR = 5 cm 2 /g of gas is adopted (Thompson et al 2015;Ishibashi & Fabian 2015). τ * ,w describes the contribution of stellar winds and supernovae remnants, for which we adopt τ * ,w = 2.5 based on Starburst99 (Leitherer et al 1999;Veilleux et al 2005;Heckman et al 2015). f c (r) is the covering factor of the starburst feedback by the outflowing OH, for which we adopt two approaches: (i) a "mixed" shell-cloud approach, in which a shell with f c = 1 is used for r ≤ R 1 (see eq.…”

“…(20), showing that the observed values are above the limiting values for most high-velocity components. By now including the gravitational term, we obtain the critical momentum fluxṖ crit (r) = M out v 2 cir /r that must be overcome at some position r byṖ * (r)+Ṗ AGN (r) in order to allow the outflow launching to proceed (Heckman et al 2015;Thompson et al 2015). In the shell-cloud approach, this condition should be met at least at r = R 1 , where (Ṗ * (r) +Ṗ AGN (r))/Ṗ crit (r) is highest, and thus we compare in Fig.…”